Please don't not break

Posted Dec 1, 2005 11:11 UTC (Thu) by smurf (subscriber, #17840)
Parent article: Future Driver core changes

>> Hopefully this will prevent any userspace program that is relying on the current structure of sysfs to not break.

Ummm... yeah. ;-)

to post comments

Please don't not break

Posted Dec 1, 2005 16:57 UTC (Thu) by gregkh (subscriber, #8) [Link] (3 responses)

Well, that's the goal. Do you know of programs relying on the current sysfs class structure that would break with these proposed changes?

Please don't not break

Posted Dec 1, 2005 17:31 UTC (Thu) by smurf (subscriber, #17840) [Link] (2 responses)

You are missing the keyword "double negation". ;-)

Read literally, your intention seems to be to have *everything* break.

Please don't not break

Posted Dec 1, 2005 17:38 UTC (Thu) by gregkh (subscriber, #8) [Link]

Doh!

Ok, sorry, I didn't mean that, thanks for pointing it out to me.

Please don't not break

Posted Jan 12, 2006 20:10 UTC (Thu) by k8to (guest, #15413) [Link]

Just to be a nudge: double negation in english is ambiguous. It is commonly used in a manner interpreted to be nonconflicting, and this nonconflicting behavior is considered proper in most languages.

All I'm saying here is humans are not CPUs, and can handle these issues fine, but it doesn't hurt to tighten up your language.

Please don't not break

Posted Dec 1, 2005 21:46 UTC (Thu) by cook (subscriber, #4) [Link]

Sorry, that slipped by me in the editing pass, the text has been fixed.

Please don't not break

Posted Jan 29, 2006 9:12 UTC (Sun) by landley (guest, #6789) [Link]

Actually, it'll break mdev in busybox. We were traversing the block and
class directory hierarchies to find dev nodes, but were avoiding
traversing symlinks because following device and bridge can lead to
endless loops.

Now we apparently have to know _which_ symlinks to follow, which sucks
deeply.